Rigol DHO814 - capturing CAN frames

[eTobey]

Hi,

i try to capture CAN frames in a row with it. I did not have any luck to achieve this in a usable manner. Always some frames will get dropped or are hidden at some time.

I tried the record function, but i cant get the intervals to match the frame. The lowest i achieved was about 3ms.
Also having like 200ms wont work, since the list only displays those frames that are visible, and only if its not zoomed to far out. Its a zoom and scroll nightmare!

[Martin72]

Come to the dark side of the force, we have cookies..

[eTobey]

Come to the dark side of the force, we have cookies..

Allright then!

[voltsandjolts]

Logic analyser + sigrok PulseView = happy life 

[eTobey]

I just joined the dark side of the force today , but i cant find how to get the list to show. Cant find it in the manual either.

[pdenisowski]

Come to the dark side of the force, we have cookies..

Or if you want to go full dark side ....

[pdenisowski]

Also having like 200ms wont work, since the list only displays those frames that are visible, and only if its not zoomed to far out. Its a zoom and scroll nightmare!

Yeah, the real trick in decoding protocol (especially CAN) is (a) making sure you have the time/div setting such that the frame(s) you're interested in fit completely on the screen and (b) triggering on the frame(s) your are interested in.

And an OBDII breakout box is an easy way to get an unlimited supply of real CAN frames to test against (see Insta post above).  Even though CAN is differentially signaled, many scopes can decode just by looking at CAN_H or CAN_L.

[nctnico]

Setting the memory size to the maximum (fixed) helps a lot when doing protocol decoding. IF the oscilloscope supports decoding the full memory (which Rigol and Siglent don't AFAIK; these decode only what is on screen. Hence the problems the OP is having). Another option is to enable zoom mode to force acquiring more data than fits on the regular screen at the cost of screen real-estate.

[Martin72]

which Rigol and Siglent don't AFAIK

Forgotten again, eh?

[Martin72]

I just joined the dark side of the force today , but i cant find how to get the list to show. Cant find it in the manual either.

Click on result list..

[nctnico]

which Rigol and Siglent don't AFAIK

Forgotten again, eh?

Well, try your test again and see if the list with decoded messages stays the same when you zoom in OR that the list contents is limited to what is on screen. As an extra test, try to zoom in on a detail of a message with the start of the message being off-screen.

[eTobey]

Well, try your test again and see if the list with decoded messages stays the same when you zoom in OR that the list contents is limited to what is on screen. As an extra test, try to zoom in on a detail of a message with the start of the message being off-screen.

Yes, it may not do the whole buffer, but it does so when zoomed pretty far out. The Rigol in this case is just useless, i mean really useless.

But there is one thing with that siglent: I cant find another way to scroll through the list. Hitting those scrollbuttons is not exactly easy! Is there a way to use a knob?
Never mind, saw a video someone using a mouse. 

[Martin72]

Yes, it may not do the whole buffer

The number of decoded values/packets is limited, so not all of the memory is used (if there is any "left over").
In the data sheets of the 2000X and 2000X HD series, the maximum number of the respective decoder could be found, in the data sheet of the 800X HD I did not find anything about it.

[Martin72]

@nctnico:
https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/msg5164788/#msg5164788

OK, it was a little bit ago...

[nctnico]

@nctnico:
https://www.eevblog.com/forum/testgear/rigol-dho804-test-and-compare-thread/msg5164788/#msg5164788

OK, it was a little bit ago...

5 seconds into the video from that posting, you can see the decoded data dissapears when the start of the packet goes off-screen. The same for the list. What isn't on screen, is not in the list with packets. Very annoying if you want to relate packets to eachother.

[eTobey]

Very annoying if you want to relate packets to eachother.

With the Rigol, you can see (and read) only one frame at once completely, and thats it. With the signlent its like hundrets? Thats quite an advantage there!

[bobxyz]

I've been playing around with I2C decoding and I think it should be similar to CANbus.  Here's what I do:

First, set the scope mode so that pressing the horizontal timescale knob switches to Zoom mode (instead of fine adjust).  Settings -> Utility -> Setup : Fine/Zoom=Zoom


Next, setup Settings -> Decode : I2C, and turn on Bus Status


Now, hit run and adjust the horizontal time base to capture as much as possible while still seeing something in the decoded protocol area.   Note: if you go slower than 20ms/div (i.e. scrolling display), protocol decode is disabled and you'll need to go back to the previous step to turn Bus Status back on.  Hit the stop button when you have a reasonable capture.  In my case, I have two I2C bursts over 100ms (the code is running 10x per second I2C polling rate).


Press the horizontal time scale knob to switch into Zoom mode, then use the horizontal position knob and the time scale knob to display a burst of I2C activity


To view the earlier I2C burst (on the left of the capture), zoom back out using the time scale knob, then using the touchscreen, drag your finger on the zoomed display to shift it back to the first I2C burst.  Zoom in and drag as needed to view the earlier I2C burst


Let me know if this isn't clear enough, and more importantly, does this work for CANbus too?

-Bob

[eTobey]

Let me know if this isn't clear enough, and more importantly, does this work for CANbus too?

I dont know, because I joined the dark side of the force already (i already sent it back).

[bobxyz]

I went ahead and tried CAN bus decoding with a DHO814 (v1.02 firmware) and it works about the same as I2C -- you need to do a big capture then zoom in (split screen) to see the decoded values.


One bug I ran across is that the CAN decode doesn't work when using a 10x probe (with a 10x scope setting).  It works with a 1x probe.  It also works with a 10x probe when the scope is left on 1x and voltage levels adjusted to 1/10th (~0.3v CAN threshold).  The CAN triggering, versus the decode display, seems to work fine for both 1x and 10x.

The screen gets rather cluttered when displaying the decode list and zoomed waveforms together.


Note: CAN decode/triggering on the DHO800 series seems to have been added in January with the later firmware updates - it wasn't supported originally, but is supported now, and is included in the newer datasheet/user manual.  LIN bus is not [yet?] supported.

P.S. How are the cookies?